I

To gain insights into how policymakers in emerging economies absorbed economic concepts and theories, a comprehensive process is required. This involves analysing these concepts and theories in their original meanings while also taking into account the specific historical contexts of the policymakers assimilating them. The objective of this study is to analyze the assimilation of the concept of ‘central bank’ by the Brazilian community of policymakers at the beginning of the twentieth century.Footnote ¹ The main historical background of the study is the debate surrounding the transformation of the Bank of Brazil into a central bank in 1923.Footnote ² At the time, the Bank of Brazil was the most important private bank in the country, and the largest commercial bank in Latin America (Marichal Reference Marichal2008, p. 587). The bank underwent a transformation in 1923, but the reform was reversed in 1926. This article endeavours to address a fundamental question: at that time what was the prevailing understanding of the term ‘central bank’ among Brazilian individuals involved in the political decision-making process, and what were the perceived functions attributed to this institution? The article aims to shed light on the different groups that advocated for and opposed the reform, delving into the nature of their support and opposition.

The analysis of Brazil is particularly interesting given its uniqueness: it is widely recognized that Brazil was one of the last independent countries in the world to formally establish a central bank. The institution that exists today in Brazil was created only in 1964. Many other countries, particularly in Latin America and other developing regions, had already established their central institution in the early twentieth century, mainly as a result of foreign missions. These countries include Peru (1922), Colombia (1923), Chile (1925), Mexico (1925), Bolivia (1928) and Ecuador (1926). Another characteristic of Brazil is that even after the establishment of a central bank in 1964, the Bank of Brazil maintained a peculiar relationship with the Central Bank of Brazil, functioning as a kind of parallel monetary authority until 1986.

Since Brazil created a central institution without foreign supervision, this article contributes to the literature that illustrates the range of scenarios leading to the establishment of a lender of last resort. For example, Moen and Rodgers (Reference Moen and Rodgers2022) show how lender of last resort operations could emerge through independent organizations, using J. P. Morgan’s actions during the 1907 crisis as an example, and Gorton and Tallman (Reference Gorton and Tallman2016) explore the role of private bank clearing houses during American financial crises prior to the establishment of the Federal Reserve. Flores-Zendejas and Marichal (Reference Flores-Zendejas and Marichal2024) argue that the emergence of central banks in Latin America during the 1920s was shaped by specific political, economic and financial circumstances, although parallels can be drawn between different historical experiences.

The Brazilian case further illustrates the ongoing debate over the role of central banks in financial crises. While one justification for establishing the institution was the country’s purported liquidity crisis, there was considerable internal disagreement not only on the appropriate measures to address it but also on whether such a crisis truly existed.

The history of Brazil’s central bank is typically focused on the creation of the Superintendency of Currency and Credit (Sumoc) in 1945 and the subsequent establishment of the central bank in 1964.Footnote ³ This article aims to expand this historiography by exploring the debates that occurred decades before these significant events. As the 1920s marked the establishment of central banks in many Latin American countries, this study also aims to highlight the distinct path of Brazil’s monetary history compared to that of other Latin American nations.

The article is primarily based on the examination of the proceedings of the National Congress and editions of the major financial newspapers, Jornal do Commercio and its Retrospecto Commercial and the Correio Paulistano. Footnote ⁴

II

In July 1917, Edwin Kemmerer, the renowned American ‘money doctor’, was invited by Venustiano Carranza, the President of Mexico, for a six-week visit. During this time, Kemmerer, among other tasks, drafted a plan for the establishment of a central bank in Mexico. Kemmerer’s subsequent missions led to the establishment of central banks in Colombia (1923), Chile (1925), Ecuador (1926) and Bolivia (1928) (Flores-Zendejas and Nodari Reference Flores-Zendejas, Nodari, Eichengreen and Kakridis2023, p. 328), and the reform of the central bank in Peru (1931) (Flores-Zendejas and Marichal Reference Flores-Zendejas and Marichal2024, p. 223). As in many other countries (Goodhart Reference Goodhart2011, p. 136), in Latin America, the creation of a central bank was primarily intended to facilitate foreign loans and investments (Drake Reference Drake1989, p. 4; James Reference James, Eichengreen and Kakridis2023, p. 49).

In Latin American countries that established central banks in the 1920s, these institutions were also designed to address monetary instability and were typically followed by the adoption of the gold standard (Flores-Zendejas and Nodari Reference Flores-Zendejas, Nodari, Eichengreen and Kakridis2023, p. 320). Singleton (Reference Singleton2011, p. 12) and Eichengreen and Temin (Reference Eichengreen and Temin2000) highlight the importance of maintaining the gold standard as a key feature of central banks during that era. Kemmerer was a staunch defender of the system and even proposed a pan-American monetary unit based on the dollar and pegged to gold (Caldentey and Vernengo Reference Caldentey, Vernengo, Battilossi, Cassis and Kazuhiko2020), and he continued to advocate for the gold standard even after the Great Depression (Drake Reference Drake1989, p. 9).

Kemmerer’s visit to Mexico had little resonance in Brazil – there is almost no mention of the mission in Brazilian newspapers.Footnote ⁵ Brazil and Argentina were the only Latin American countries between the 1890s and 1920s that did not employ American financial advisors (Drake Reference Drake1997, p. 94).

Yet, in 1917, there was a discussion in the country about the need to create a lender of last resort. Homero Baptista, then president of the Bank of Brazil, suggested in his annual report the transformation of the institution into a central bank. According to Baptista, the bank could become a favorable factor in the development of the economy if structured similarly to its counterparts in England, France and Germany (Pacheco Reference Pacheco1979, p. 120). Baptista studied at the Law School of São Paulo.

Notably, as early as the 1830s, directors of the Bank of England debated whether the institution should prioritize public interests (Bignon et al. Reference Bignon, Flandreau and Ugolini2012, p. 582). In this context, it is significant that the president of the Bank of Brazil explicitly envisioned the bank as an instrument for national development.

In December 1918, Rafael de Abreu Sampaio Vidal presented the bill that served as the foundation for the Chamber’s discussions for the following years and the basis for the transformation of the Bank of Brazil in 1923. The bill presented by Vidal in 1918 aimed to establish a Central Bank of Issuance and Discounts, named Bank of Brazil, headquartered in Rio de Janeiro (National Congress [henceforth NC], 11 Dec. 1918). Sampaio Vidal had also graduated from the Law School in São Paulo, owned a coffee plantation and founded the Brazilian Rural Society. He was elected by the Paulista Republican Party (PRP), the most influential political party in São Paulo during the Old Republic (1889–1930), and was a staunch advocate for maintaining coffee prices in international markets. At the time of the transformation of the Bank, in 1923, Sampaio Vidal was the Minister of Finance of Brazil.

Sampaio Vidal’s proposal envisioned the creation of an institution headed by a president appointed by the federal government and a board of eight members: three directors representing national banks and five directors elected by shareholders. However, the representatives of national banks were to have only advisory roles (p. 4849). Despite this, the proposal advocated for private banks to have priority in subscribing to the new shares. The federal government would be prohibited from holding shares in the bank. That differed, for example, from Mexico, where the government held 98 percent of the shares of the central bank (Marichal Reference Marichal, Dobado, Galvarriato and Márquez2007).

It was proposed that the bank should have the exclusive privilege to issue notes, with an issuance limit of 600,000 contos of réis, backed by 200,000 contos of réis in gold and the remainder in various financial instruments.Footnote ⁶ These were the ‘commercial effects’, a very broad category of securities that include, for example, debt securities for coffee, rubber, frozen meats and other goods.

Sampaio Vidal argued that the limit of monetary circulation should be based on the real needs of circulation. The limit for monetary expansion should be ‘real business and not dangerous speculation’ (p. 4849). Backing in ‘real’ assets was considered even better than gold backing (p. 4850).

For Sampaio Vidal, a central bank functioned as a rediscount bank, meaning it served as a lender of last resort (p. 4844). However, it was also part of his plan that the central bank should operate as a discount bank, defining such a bank as one that relies solely on its capital and depositors’ funds. The bank would not only act as a lender of last resort but would also operate like a commercial bank, discounting commercial and industrial bills.

Vidal attributed recurring money shortages to the lack of support for commercial banks. He described frequent and abrupt financial crises, where even minor market fluctuations led banks to lock their vaults, deny credit and push clients toward bankruptcy – refusing even the most secure transactions backed by valid guarantees (NC, 11 Dec. 1918, p. 4844).

He advocated for a ‘resilient institution’ capable of maintaining economic stability in normal times and providing support during crises (p. 4844). Other senators observed that banks were stockpiling reserves out of fear of bank runs, underscoring the need for a central institution to provide liquidity when needed (NC, 25 Dec. 1918, p. 5414).

The concept of the central bank as the bank of banks is part of the classical theory of central banks, primarily developed by Henry Thornton in the nineteenth century (Rist Reference Rist1940, p. 382). Worldwide, the lender of last resort function, often regarded as a central bank’s primary role, grew increasingly important in the late nineteenth century (Bordo and Siklos Reference Bordo and Siklos2017, p. 4). However, the period following World War I represented a confusing interregnum, marked by considerable uncertainty regarding the proper role of central banks during the 1920s compared to earlier years (Goodhart Reference Goodhart2011, p. 136).

The issue of circulation elasticity is repeatedly emphasized: the money supply should align with the volume of business, expanding during periods of growth and contracting during downturns. It was necessary not only to have resources to develop production but also to stabilize product prices in the market, guaranteeing contracts at predetermined prices (p. 4847). Given that Sampaio Vidal and many of the bank’s supporters were coffee growers, it is important to highlight their practice of collectively withholding production – which required large-scale financing – with the aim of manipulating coffee prices in the international market, as well as the recurring disagreements across different levels of government on how to fund this stock retention.

The accumulation of gold reserves without using them to increase the money supply was severely criticized. Later, Sampaio Vidal argued: ‘What fundamental difference is there between the gold from the depths of the earth and the gold from the depths of the safes? What difference is there between this useless gold and the golden calf that was worshipped in ancient times?’ (Correio Paulistano [henceforth CP], 23 Dec. 1922, p. 5).

The bank could also, ‘in case of extreme necessity’, issue beyond the 600,000 contos of réis, provided it paid an 8 percent tax to the Treasury on the excess issuance. The project also proposed the creation of a special portfolio, endowed with 200,000 contos of réis – one-third of the bank’s capital – for extraordinary operations to support agriculture and pastoral activities. This measure was modeled after the example of Austria, which had established a special portfolio in its issuing bank to support agriculture, functioning since 1855: ‘We would not, therefore, be creating in Brazil an institution never seen, never tried’ (NC, 11 Dec. 1918, p. 4849).

Sampaio Vidal outlined the functions of a central bank: first and foremost, it is the regulator of monetary circulation (p. 4045). Additionally, it should ‘address and serve all the needs of the national economy, the productive and commercial classes … ensuring the development of the country’s agricultural, pastoral, industrial, and commercial sectors’ (p. 4045). He added that the bank was ‘the supreme guarantee of economic expansion’ (p. 4045). In other words, the central bank should aim not only to regulate the monetary system but also to promote the country’s development.

Vidal categorizes the world’s central banks into three main types. The Bank of England operated under the strictest rules, requiring that banknotes issued had to match the exact amount of gold reserves. Due to this rigidity, the bank lacked the necessary elasticity demanded by the economy, according to him (p. 4645).

The second type of central bank was represented by the Bank of France. The state set a maximum limit on the bank’s issuance of notes, which could be adjusted according to economic needs. There was no legal requirement for gold backing, leaving it to the discretion of the Bank of France’s management. This flexibility, according to Sampaio, allowed the Bank of France to contribute to the country’s full economic development.

The German Reichsbank, the third type, had an even more elastic money supply, according to Sampaio, maintaining one-third of its reserves in gold, with the remainder in financial instruments. If the bank exceeded its maximum allowed issuance, it had to pay a 5 percent tax to the state. He concluded that the Reichsbank was the most advanced and best suited to the needs of the modern world (p. 4846).

Vidal quoted famous defenders of the real bills doctrine, Bernardo de Souza Franco and Rui Barbosa (p. 4847) – the latter being notorious for the failed experiment with multiple bank issuers at the end of the nineteenth century, which led to rampant inflation in Brazil. Rui Barbosa’s role is repeatedly praised, with Vidal referring to him as ‘the greatest financial mind of our nation’ (p. 4848).

Commenting on the situation in the United States during the nineteenth century, Sampaio observed that the country adopted the worst possible system – one of plurality in currency issuance (p. 4846). The greatest danger was the low elasticity of the money supply, posing a constant risk of collapse during crises (p. 4846). He also criticized the backing of currency with public debt securities (p. 4846).Footnote ⁷

Regarding the state involvement, Sampaio was clear: ‘let us not create a state bank’, emphasizing that while the government should oversee the activities, the management should remain with the shareholders (p. 4847). He argued that the bank should be a private institution, with the state regulating its establishment and operations through legislation and acting as a supervisor (p. 4848). Criticizing the issuance of money by the government rather than a central bank, Sampaio questions: ‘Who oversees and regulates the government?’ (p. 4847). Vidal further critiques the government’s management of foreign currency reserves, pointing to what he sees as the erratic practice of some ministers using these funds to attempt exchange rate appreciation, throwing reserves acquired through sacrifice into the ‘blazing furnace of currency speculation’ (p. 4847).

For Sampaio, the central bank also regulates circulation, avoiding the inconveniences of inflation. To achieve this, it should use discount restrictions and raise interest rates. The issue of price stability was a common justification for the creation of central banks in other Latin American countries. However, in some cases, the emphasis differed; for example, in Argentina, discussions about creating a central bank had been ongoing since the late nineteenth century, but the focus was on maintaining stable prices to preserve the purchasing power of wages (Flores-Zendejas and Nodari Reference Flores-Zendejas, Nodari, Eichengreen and Kakridis2023, p. 328). In the years following the war, some economists began to advocate that the central bank’s role should include price stabilization. In the United States, a Senate committee was formed in 1928 to examine whether the Federal Reserve should adopt price stabilization as part of its official mandate (Rist Reference Rist1940, pp. 425–6).

The central bank’s ability to set interest rates is also emphasized. For Vidal, this was a crucial channel through which the bank could stimulate national production. Other politicians supported this idea, noting that banks had resources but were hesitant to lend due to fears of bank runs (NC, 25 Dec. 1918, p. 5413). If there was a lender of last resort – through rediscounting – banks would be more willing to lend, leading to lower interest rates across different regions of the country (p. 5413).

A significant debate arose: should the existing Bank of Brazil, reorganized in 1906, be used, or should a new institution be established? Vidal argued that if the Bank of Brazil were to be used, three conditions must be met: (a) reduce the bank’s capital based on asset evaluation, accounting for losses or illiquid assets; (b) prioritize national banks in the subscription of new shares; and (c) reconstitute the bank according to joint-stock company laws (NC, 23 Dec. 1918, p. 5416).

Since the nineteenth century, papelistas, like the metalistas, regarded currency convertibility as the standard for normal monetary organization (Villela Reference Villela, Bielschowsky, Boianovsky and Coutinho2017, p. 115).Footnote ⁸ Even leading papelistas like Barão de Mauá viewed fiat money as a temporary measure (p. 116). Vidal also saw the proposed arrangement as temporary until Brazil could adopt currency convertibility. He believed the country needed time to prosper and retain gold effectively, arguing that this was the natural approach, rather than imposing convertibility prematurely (NC, 11 Dec. 1918, p. 4848).

One senator from Minas Gerais, in southeastern Brazil, João Luiz Alves, agreed, arguing that convertibility should wait until the balance of payments stabilized and external debt was reduced. He emphasized developing domestic industry, including manufacturing and agriculture, to reduce imports and improve the trade balance, warning of the urgent risk of foreign products flooding the domestic market after the war (NC, 25 Dec. 1918, p. 5412).

The central bank was viewed as essential for developing an internal system of productive financing, especially given the lack of external capital (NC, 11 Dec. 1918, p. 4849). Vidal argued that reliance on foreign capital was not advisable, particularly with the erratic exchange rate. Even if international loans were available, global instability made borrowing too risky. For Sampaio, this was about national sovereignty: the country needed to address its challenges using its own resources and strengths (p. 4849).

By the end of 1918, the specialized press also seemed to support the creation of a central institution, with Bank of Brazil being the most obvious candidate: ‘Bank of Brazil is destined to play this role, and it seems that the most viable solution would be its transformation’ (Retrospecto Commercial [henceforth RC] 1918, p. 60). However, Brazil should have the courage to adapt theories to its reality, just as the United States built the Federal Reserve by drawing on European doctrine but tailoring it to American conditions (RC 1918, p. 60).

The president elected in 1919, Epitácio Pessoa, invited Homero Baptista, the former president of the Bank of Brazil who had already brought the issue of transformation to the forefront in his 1917 report, to serve as the Minister of Finance in his government.

III

Despite Sampaio Vidal’s project, the urgency of the situation led to the debate to create a special Rediscount Office in the Bank of Brazil to deal with the shortage of money supply. Sampaio Vidal, however, raised concerns about the creation of the Office within the Bank of Brazil. He argued that the proposed amount of 100,000 contos of réis was insufficient given the significant business expansion in recent years; it would barely suffice for the needs of Rio de Janeiro alone (NC, 8 Oct. 1919).

At that time, Washington Luís, then the governor of São Paulo and later President of Brazil in 1926, was another advocate for creating a central bank in Brazil. He supported the initiative, believing it was crucial to securing financial assistance from the Rothschild House (Novelli Reference Novelli2002). In his administration platform, presented on 25 January 1920, Washington Luís compared a strong banking system to the circulatory system of the human body, arguing that without it, economic life would be impossible (CP, 28 Jan. 1920, p. 1).

In early August 1920, the Commercial Association requested a meeting with the finance minister, urging the transformation of the Bank of Brazil into an issuing bank (RC 1920, p. 9). The representation argued that if an issuing bank had existed, there would have been no crisis. They claimed that if the ability to retain and secure production through warranting (warrantagem) had been available, any attempts by buyers to acquire those goods at low prices would have been rendered futile (RC 1920, p. 10).

The timing of the commercial association’s request was significant. Money shortages typically worsened from August onward due to the harvest season, and debates over the creation of a central bank always intensified in the latter half of the year. This issue was not unique to Brazil; in the United States during the post-Civil War era, many pointed out that, in the absence of a central bank, the system could not expand credit quickly enough during the harvest season (Gorton and Tallman Reference Gorton and Tallman2018, pp. 39-40).

One senator from São Paulo, Alfredo Ellis, a physician trained at the University of Pennsylvania and also a coffee plantation owner, emphasized that the creation of a rediscount bank was an urgent measure (NC, 25 Aug. 1920, p. 2360). According to the senator, a rediscount bank would enable legitimate, non-speculative commerce to have the necessary resources to protect its earnings (p. 2361). However, a deputy from Piauí in northeastern Brazil – far from the coffee producers of São Paulo – Armando Burlamaqui, argued that defending coffee prices was a private matter for the coffee growers, not a concern of the Brazilian nation.

On 3 September, Carlos de Campos, a deputy from São Paulo, introduced a bill proposing new conditions for the issuance of paper money to provide loans to agriculture and industry (RC 1920, p. 14). However, the discussion of the bill was postponed due to the arrival of King Albert of Belgium (RC 1920, p. 15).

King Albert I of Belgium and his wife, Queen Elisabeth, visited Brazil from 19 September to 15 October 1920. The mission is still regarded as one of the most lavish state visits in Brazilian diplomatic history. The Brazilian press highlighted it as the first visit by a European monarch to a republic in South America (A Noite, 2 June 1920, p. 1). However, in Congress, there was strong criticism of the extravagant expenses associated with the visit, which reinforced the idea that the Government – particularly the Treasury, then responsible for issuing money – was not reliable enough to be responsible for managing such important matters as the money supply.

Sampaio Vidal, who initially opposed the creation of the Rediscount Office, later regarded it as an intermediate solution (NC, 30 Sept. 1920, p. 3202). However, Carlos Maximiliano, a deputy from Rio Grande do Sul, south Brazil, voiced his dissent. He argued that it could not serve as an intermediate solution, as there were no restrictions on either the duration or the amount of paper issued (p. 3202). He further contended that, after 14 months of deliberation, the country had finally reached a consensus – a rare occurrence in Latin America, according to him – that the establishment of a central rediscount bank was the appropriate solution. The portfolio, in his view, directly contradicted the conclusions of these discussions (p. 3202).

In a broader defence of bank-issued currency – as opposed to government-issued currency, that is, Treasury notes – Maximiliano cites Wilhelm Roscher and his book Nationalökonomik des Handels und Gewerbfleißes: ‘Of all the candidates for money, the best is a good banknote’ (p. 3204). Government-issued money, he argued, lacked the ability to adjust to market conditions with the necessary agility (p. 3204). Maximiliano had also graduated in Brazil, from the Law School of Belo Horizonte.

Maximiliano also identified the source of the country’s aversion to the creation of a central bank: the mistakes made by the Provisional Government (1889–91) (p. 3204). As previously noted, the early years of the republic were marked by the excessive issuance during the Encilhamento, followed by a crisis that eventually led to the liquidation of the Bank of the Republic of Brazil (Franco Reference Franco and Abreu2014, p. 44). Furthermore, Brazil had experienced other earlier attempts at establishing issuing banks, all involving institutions named ‘Bank of Brazil’, with bad results. For instance, the first Bank of Brazil, founded in 1808, was liquidated in 1829, and the free banking experiments in the nineteenth century that saw two versions of the Bank of Brazil also ended in failure (Franco Reference Franco2017, p. 277). The press in Brazil echoed this perspective, attributing the failures to unfavorable circumstances at the time these banks were established and to abuses in credit and Treasury withdrawals (RC 1922, p. 29).

In 1920, the first international conference of the League of Nations was held in Brussels, attended by figures such as John Maynard Keynes, Arthur Pigou and Gustav Cassel. There was advocacy for nations without central banks to establish them. Central banks of issuance were to be free from political pressures and managed according to prudent financial principles (James Reference James, Eichengreen and Kakridis2023, p. 48). The reconstruction of the gold standard was a central topic of debate (Clavin Reference Clavin, Eichengreen and Kakridis2023, p. 67). This issue was also a concern for the new governor of the Bank of England, Montagu Morgan, who took office in 1920 and played a prominent role in the Brussels conference, advocating for cooperation among central banks to ensure the restoration of the gold standard (Clement Reference Clement, Eichengreen and Kakridis2023, p. 82). The recommendations from the Brussels conference were viewed in the Brazilian press as aligned with the ‘laws of the classical liberal school’ (RC 1920, p. 6). For ‘renowned economists’, the main goal was to provoke deflation and then stabilize production and consumption (RC 1920, p. 6). The Brazilian press noted that the conference suggested avoiding the artificial maintenance of high discount rates, recommending instead the use of the natural rate (RC 1920, p. 7).

Antonio Carlos Ribeiro de Andrada, a graduate of the São Paulo Law School and a deputy representing Minas Gerais, was the principal opponent of the project to transform the Bank of Brazil. He also highlighted – citing Cassel and Keynes – that the main recommendation of the Brussels conference was inflation control (Andrada Reference Andrada1923, p. 470). Ribeiro de Andrada, during one of the discussions on the currency issuance project, opposed the emergency issuance, arguing that even at high interest rates, it was preferable to attempt borrowing from abroad (NC, 17 Oct. 1920, p. 3572). According to Andrada, ‘our production is not in crisis, and there is no reason to foresee one’ (NC, 17 Oct. 1920, p. 3572).

The lack of consensus on the existence of a crisis is significant. Although it is often controversial to pinpoint when society recognizes a crisis, the 1907 crisis, which led to the creation of the Federal Reserve, for example, was a clear instance of financial distress, signaled by significant events such as the suspension of deposit-to-cash conversions by the New York Clearing House (Gorton and Tallman Reference Gorton and Tallman2018, p. 38). This is important because, as the literature shows, the forms taken by the role of a lender of last resort are heavily influenced by how economic models portray the emergence of financial panics (Bignon et al. Reference Bignon, Flandreau and Ugolini2012, p. 584). Bordo and Siklos (Reference Bordo and Siklos2017, p. 4) find that most major central banks before the twentieth century were created for fiscal needs (such as war financing) or to curb financial crises.

Ribeiro de Andrada pointed out that, since the war, there had been the formation of a strong and persistent inflationary mentality, which continued to endure (Andrada Reference Andrada1923, p. 362). He argued, based on the lessons of political economy, that banks should not be allowed the right to issue paper money; this right belongs exclusively to the State. Banks should only be granted the right to issue convertible money (pp. 398–9). In his view, it was only political amnesia, which had forgotten the traumatic experiences of the nineteenth century with issuing banks, that was responsible for the continued insistence on this practice (p. 399). Andrada defended that central banks should discipline the money supply, but ‘instead of seeking, like their predecessors, to sanitize the circulating medium, the new plan for banknote issuance is more likely to corrupt and distort it’ (p. 401).

Ribeiro de Andrada argued that financing coffee purchases by issuing paper money amounted to socializing agricultural losses (NC, 17 Oct., p. 3572). Citing Russian economist Arthur Raffalovich, he described inflationism as a ‘state of psychopathy’, often leading to extravagant issuance projects (p. 3572). Ribeiro de Andrada warned that forced money issuances are destructive, regardless of their form, even when through rediscount offices or issuance banks, which he called ‘paper money factories’ (p. 3575). He proposed that if government expenses exceeded revenues, the government should be allowed to secure foreign loans up to 30,000 contos of réis (p. 3575).

Amendments were made to the proposed money issuance law, including especially one by Bento Miranda, a deputy from Pará, north Brazil, proposing the creation of a rediscount office at the Bank of Brazil and allowing the government to issue currency backed by commodity guarantees (p. 3576). Bento Miranda then cited Keynes, emphasizing that excessive importance was placed on the correspondence between currency and gold, while insufficient attention was given to the broader development of the banking system, for example the use of bank checks (p. 3580).

It is worth noting that, despite the frequent references to Keynes, and also to Cassel, it was not possible to find, in the discussions of Brazilian politicians, significant mentions of the literature more commonly associated with central banks, such as the works of Walter Bagehot. Stanley Jevons, on the other hand, appears as a reference in monetary theory. Retrospecto Commercial (1920, p. 8; 1922, p. 38) discusses Jevons’ ideas regarding clearing houses, for example. Bagehot articulated fundamental principles for central bank operations, recommending that they provide liquidity during crises by lending against sound collateral and at high interest rates.Footnote ⁹ In Brazil, however, besides not referencing Bagehot’s ideas, the debate also did not include discussions on well-defined guidelines for central bank operations during a crisis.

For creditor countries, according to Bento Miranda, the defence mechanism should function like a sluice gate, preventing the outflow of gold. This was the role of the issuing banks. For debtor countries, the appropriate defence mechanism was conversion offices and reserve funds, which supplied gold in times of scarcity and accumulated it in times of abundance. Hence, he proposed the creation of a Rediscount Office of Bank of Brazil. In summary, although many saw it as an intermediary step toward establishing a central bank, Bento Miranda, who proposed the creation of the rediscount facility, believed that central banks were not suitable for Brazil.

Bento Miranda categorized issuing banks into three types: the Bank of England, the Bank of France, and the Reichsbank. Germany was a hybrid country, sometimes a creditor and sometimes a debtor. Like Germany, there were the ‘second-rate countries’ – nations that were alternately creditors and debtors, such as Belgium, the Netherlands and Greece (p. 3581). Unlike Brazil, however, these countries had access to the international short-term loan markets. Even if Brazil accumulated enough gold to cover one-third of its issuances, it still lacked the high-quality short-term securities to cover the remaining two-thirds. Just because institutions worked well in one country did not mean they would necessarily work in Brazil (p. 3582). He noted that although the National Congress had passed legislation on the use of checks, they were still not widely adopted by the population.Footnote ¹⁰ In the United States, for example, during the National Banking Era (1864–1912), prior to the creation of the Federal Reserve, certified checks were commonly used as a mean of payment (Gorton and Tallman Reference Gorton and Tallman2018, p. 91).

Bento Miranda justifies his position by drawing on mechanical principles: he explains that certain machines, such as the triple-expansion engine, perform optimally when there is an ample supply of water, which is why they are commonly used in maritime activities. Without enough water, their efficiency drops significantly, though an untrained observer might think the machine is working fine, unaware of its poor performance: ‘This is what is happening with the Reichsbank – the mechanism is not functioning properly. It issues depreciated paper currency, discounts depreciated securities, but lacks the power to defend the collapsing mark’ (p. 3582).

Sampaio Vidal disagreed, arguing that the Reichsbank was so well structured that it provided significant benefits to the national economy even during the war (p. 3582). However, Bento Miranda countered: ‘The model is now crippled; we should not start walking by imitating a limping entity’ (p. 3582). He believed Brazil should only establish a central bank once it transitioned from a debtor to a creditor country. Raul Alves, a deputy from the Brazilian northeast, opposed this view, pointing out that Japan began its economic development with issuing banks.

Bento Miranda noted that monetary ‘elasticity’ meant expanding the money supply during economic growth and contracting it during recessions. However, he argued this pro-cyclical approach would actually worsen economic crises (p. 3582).

Bento Miranda noted that some preferred issuing currency backed by assets, even if just securities, over issuing it without any backing. However, he disagreed, arguing that it was better to issue currency without backing, as long as gold was used to stabilize the exchange rate (p. 3585). Sampaio Vidal argued that a central bank’s role was to stabilize the foreign exchange market, but Bento Miranda countered that this strategy worked only for creditor nations, which could prevent the outflow of gold from their economies (p. 3586).

Raul Cardoso, from the Paulista Republican Party and a professor at São Paulo’s Law School, also opposed the creation of the bank. He outlined three main purposes of money: (1) settling international debts; (2) overseeing and regulating domestic exchanges; (3) saving or hoarding (p. 3582). For international exchanges, it was customary to use gold. For saving or hoarding, the most attractive options were also gold or silver. But to oversee and regulate domestic exchanges, ‘the currency the population is accustomed to works perfectly’ (p. 3582). Cardoso emphasized that settling international debts was the most important function of gold, as it upheld national sovereignty and independence. Nations with ample gold could use it for all purposes, but those with less should prioritize the most critical functions. He criticized the project to transform the Bank of Brazil, arguing that even with one-third backing, using gold for circulation risked neglecting its essential role in settling international debts. This left gold vulnerable to speculation and market forces (p. 3582).

Carlos Penafiel, a deputy from south Brazil, also opposed the creation of a central bank. Using France as an example, he noted that the Bank of France had become a ‘cashier’ for other banks, and Brazil was attempting to replicate this (p. 3691). Although he favored banking freedom, he believed that Brazil, lacking full industrial development, should keep the privilege of issuance under state control. Quoting Auguste Comte, Penafiel argued that the government could more easily give up a monopoly – the issuance monopoly – than private institutions, like a private central bank.

On 29 October 1920, the Finance Committee of the Federal Senate held an extraordinary session to discuss the issuance project and its many proposed amendments. Senator Francisco Sá, after consulting with President Epitácio Pessoa and acting as rapporteur, introduced an amendment to the Chamber of Deputies project. To expedite the urgent measures needed by productive sectors, Sá proposed creating a Rediscount Office at the Bank of Brazil. Although a special law, instead of an amendment, would have been more advisable, it would have taken too long, according to him. The amendment established an issuance and rediscount facility ‘until a special bank could be created’ (RC 1920, p. 20).

The press viewed the facility as a step toward establishing a central bank (p. 7). It was authorized to issue up to 100,000 contos of réis, primarily backed by ‘commercial effects’. Exceeding this limit required a special Act passed by the President of the Republic (p. 17).

The amendment established a unique relationship between the Bank of Brazil and the Treasury, a defining feature of Brazil’s monetary system even after the creation of the central bank in 1964. The final law, Law 4182 of 13 November 1920, titled ‘Supervision of Banks and Banking Houses’, also included provisions to curb speculation in the foreign exchange market (p. 19).

The law establishing the Rediscount Office was swiftly amended by Article 50 of the Law 4230 on 31 December 1920. Paragraph 7 further simplified issuance by allowing the Bank of Brazil’s own securities to back the Rediscount Office (Andrada Reference Andrada1923, p. 405). Although the bank’s president, José Maria Whitaker, initially took a conservative view, treating issuances as a crisis measure, the Rediscount Office became increasingly expansionist (p. 417).

In August 1922, both the Chamber and Senate approved the use of Treasury loan securities, alongside the Bank of Brazil’s own securities, as backing for paper money issuance (p. 442). Critics, such as Ribeiro de Andrada, argued it was absurd for the government to pay interest to the Bank on funds it had authorized the Bank to hold, benefiting only the shareholders through increased dividends and directors’ salaries (p. 442).

In Sampaio Vidal’s original proposal, up to 10 percent of the bank’s net profits would go to shareholders (NC, 11 Dec. 1918, p. 4852). The rest would fund a 5 percent reserve, with the remainder split equally between gold conversion to strengthen reserves and the withdrawal from circulation of Treasury-issued notes (11 Dec. 1918, p. 4852).

Neuhaus (Reference Neuhaus1975, p. 65) views the Rediscount Office as a key factor in the economic recovery by mid 1921, arguing that its success and popularity gave significant momentum to the movement for establishing a central bank.

IV

In 1920, shortly after the Rediscount Office was established, calls for the creation of a central bank continued from politicians in both the southeast and northeast, as well as the Brazilian press (NC, 27 Dec. 1920, p. 6649; NC, 19 July 1921, p. 1850; RC 1921, p. 48).

In response to widespread demand, presidential candidate Arthur Bernardes devoted a significant part of his platform to the urgent need for a central bank for issuance and rediscount (RC 1922, p. 29). He emphasized to Congress that Brazil remained one of the few civilized nations without a central bank (NC, 8 Nov. 1921, p. 6054).

The argument that fiduciary emissions were a hallmark of an ‘uncivilized’ world was a recurring theme. The concept of ‘civilization’ was frequently emphasized in Congress and public speeches, as exemplified by Sampaio Vidal during banquets (CP, 23 Dec. 1922, p. 5) and in statements claiming that the country violated all principles upheld by civilized nations (NC, 11 Dec. 1918, p. 4847).

According to Bernardes, then a presidential candidate, without a central bank, the development of industry and commerce would remain precarious (NC, 8 Nov. 1921, p. 6054). The timing was favorable, as many countries had suspended convertibility, giving Brazil time to prepare before returning to the gold standard (p. 6054). Bernardes argued that the federal government should merely oversee the bank, which would function as the bank of banks. He argued that only a bank of issue could organize agricultural credit, providing loans between planting and harvest while protecting against market pressures. Even if inconvertible, banknotes – unlike Treasury notes – would inspire greater confidence (p. 6054).

Grossman and Rockoff (Reference Grossman, Rockoff, Bordo, Eitrheim, Flandreau and Qvigastad2016, p. 232) identify two perspectives on the lender of last resort: a narrow view, which prioritizes supporting banks to sustain the money supply, and a broader view, which extends assistance to markets and institutions. Statements like Bernardes’s, which highlight the bank’s role in financing agricultural activities, suggest that the Brazilian proposal may align more closely with the broader interpretation of lender of last resort.

After being elected, Arthur Bernardes invited Sampaio Vidal – who had written the creation project in 1918 – to serve as Minister of Finance. In December, Cincinato Braga, a deputy from the same party (PRP) and also a graduate of the São Paulo Law School, proposed a budget amendment to transform the Bank of Brazil into an issuing bank (RC 1922, p. 29). Braga argued that opposition to the institution came only from enemies of capitalism, such as ‘communism, Bolshevism, and similar aberrations’ (p. 66). He was later appointed president of the Bank of Brazil.

The process moved swiftly: once Braga’s bill was introduced, it underwent only minor changes in the Senate before becoming law on 13 January 1923 (p. 29). This rapid progression exemplifies what Andrada (Reference Andrada1923, p. 362) criticized: the political customs of the country often left the executive branch with near-complete control over final and important decisions on major public interest issues. The way the central bank was established – and dismantled three years later – also supports Marichal’s (Reference Marichal, Dobado, Galvarriato and Márquez2007) argument, who contends that, unlike in Western Europe, where central banks developed gradually through an evolutionary process, the Americas experienced discontinuous shifts in central bank organization, recurrently driven by political and economic circumstances.

The Treasury was to transfer 300,000 contos of réis in gold to the Bank of Brazil, representing its debt to the Bank. In return, the Bank would gain the exclusive right to issue paper money, effectively ending the Treasury’s authority in this regard (RC 1922, p. 29).

Money issuance was to follow Sampaio Vidal’s 1918 proposal: backed by gold and commercial securities, with one-third of reserves in gold and the remaining two-thirds in credit securities from Brazil’s ‘most solid commercial firms’ (RC 1923, p. 143). However, an exception allowed the bank to increase issuance during extreme necessity, as recognized by executive decree, with a 12 percent interest rate. The law also stipulated that Brazil’s currency would become convertible to gold if the exchange rate of 12 pence per mil-réis was maintained for three consecutive years – a goal widely utopian. When asked how long this would take, Braga replied: ‘Only God knows’ (RC 1922, p. 31). In 1923, the exchange rate fell to its lowest point during the Old Republic (1889–1930) at 4 3/8 pence per mil-réis (Normano Reference Normano1939, p. 249). Braga later added that convertibility would also require the metallic reserve stock to reach 60 percent of the issuance value (RC 1923, p. 146).

Despite Sampaio Vidal’s original plan, which specified that the government would not hold shares in the bank, 1923 marked the year the federal government became the majority shareholder of the Bank of Brazil (Andrade and Deos Reference Andrade and Deos2009, p. 49). The bank’s shares were incorporated into the nation’s alienable assets.

A recurring concern at the time was the risk of the Bank of Brazil going bankrupt, especially given its organization in 1905 and the memory of the 1898 banking failures, which included the Bank of the Republic of Brazil (Franco Reference Franco and Abreu2014, p. 44). In response to this concern, Braga argued that since the Government was the Bank’s main shareholder, it could only go bankrupt after the National Treasury, the State Treasuries, the Municipal Treasuries and all wealthy men in the country had gone bankrupt (RC 1923, p. 146). He further suggested that the creation of a National Mortgage Bank, with half its profits going to the Bank of Brazil, could bolster its financial stability (p. 150).

Although many critics of central banks, in general, argue that their establishment encourages private banks to be less prudent with loans that could lead to bankruptcy, in Brazil, the concern instead – and perhaps more troubling – was that the Bank of Brazil itself might face insolvency.

Critics raised concerns about the Bank of Brazil’s dual role as a commercial bank. Jerônimo Monteiro, from the Espírito Santo Republican Party, argued that the bank would prioritize its own commercial interests when in need of funds, essentially conducting transactions with itself as both issuer and primary customer (Federal Senate 1922, p. 476). This critique reflected traditional views, such as those of Montagu Morgan, president of the Bank of England, who argued that central banks should not compete with other banks (Pringle Reference Pringle2001, p. 30), and it resonates with arguments from the nineteenth-century English discussion about the public responsibilities of the Bank of England (Bignon et al. Reference Bignon, Flandreau and Ugolini2012, p. 582).

The fact that the Bank of Brazil needed to continue operating as a commercial bank and not solely as a rediscount bank – was justified by Braga, president of the Bank, by the poor state of banking development in Brazil. In the United States, there were over 30,000 banks. In Brazil, given the low number of banks, the issuing bank needed to engage in a wide range of business activities (RC 1923, p. 147). However, it was acknowledged that ‘from a scientific principles perspective’, this was not an ideal situation (p. 147). The Bank of Brazil was responsible for one-third of the banking sector’s activities in the country during the 1920s (Baer Reference Baer2007, p. 320).

Brazil hosted a significant mission in the 1920s: Edwin Montagu, who had served as the British Financial Secretary to the Treasury, visited Brazil in 1924. He observed that the 1923 monetary reform had formally transferred control of monetary policy to the Bank of Brazil. Montagu noted that if government influence could be curtailed, the Bank could enforce monetary discipline in the country (Fritsch Reference Fritsch1980, p. 738). However, the deflationary policies of Arthur Bernardes’s government in 1926 reversed the reforms of 1923. As Fritsch (Reference Fritsch1993) notes, ‘the independent central bank created by Cincinato Braga had turned into a money-printing machine to finance the coffee industry’.

The idea of establishing a central bank in Brazil would resurface in the early 1930s with the arrival of another British mission, led by Otto Niemeyer (Abreu Reference Abreu1974, Reference Abreu1998, p. 7). During a visit to London in 1930, the then president-elect of Brazil, Julio Prestes, mentioned the possibility of employing Edwin Kemmerer during a consultation with the Rothschilds. The Rothschilds suggested instead hiring an English expert, arguing that an English specialist would not use the opportunity ‘to promote national or personal interests’ (Abreu Reference Abreu2002, p. 528).

V

The difficulties following World War I led many Latin American countries to hire foreign missions, especially from the United States, to implement structural changes in their financial systems and attract foreign loans. Brazil, however, took a somewhat different path: in the post-World War I period, some policymakers opposed – at least in discourse – borrowing money from abroad. In this context, the creation of a central bank and the development of a domestic credit market were seen as solutions to address the challenge of maintaining coffee prices – the country’s main commodity and primary source of foreign exchange. Given that the missions of ‘money doctors’, such as Kammerer’s, were typically privately contracted by governments, the scepticism of key figures in Brazilian politics – such as Sampaio Vidal, the Minister of Finance at the time – toward external assistance might partly explain why Kammerer was not hired as he had been by other Latin American governments. In Sampaio Vidal’s rhetoric, the Central Bank was viewed as a means of addressing Brazil’s problems using the country’s own resources.

The debate over the establishment of a central bank in Brazil closely followed the main monetary issues discussed during the Old Republic (1889-1930): issuing rights and convertibility. The discussions surrounding the Central Bank mirrored the controversies between metalistas and papelistas in Brazil, which also had counterparts throughout Latin America. But in Brazil, the push for creating a central bank was taken up by the group more favorable to monetary issuance. The central bank in Brazil was not primarily intended as a means to establish the gold standard, although it was considered a remote possibility.

The analysis also allows comparisons with the United States. While proponents of the bank’s creation cited similar concerns about cash shortages during harvests, Brazil – unlike the United States in 1907 – lacked consensus on the existence of a crisis. Some legislators denied a crisis, and others viewed the coffee financing issue as limited to São Paulo.

Opponents of the project, most notably Carlos Ribeiro de Andrada, were aware that a central bank should play a role in restricting – or ‘sanitizing’ – the money supply. The global uncertainty about the role of central banks after World War I, along with the fact that Brazil’s establishment of the institution was not directly influenced by a foreign mission, may have contributed to the unique way the concept of a central bank was integrated into Brazil’s economic thought at the beginning of the twentieth century.

Although the main theories about central banks at the time emphasized the importance of central banks not performing the functions of private banks to avoid competition, the Bank of Brazil continued to operate as a private bank even after its transformation in 1923. The official justification was that the Brazilian banking sector was still very underdeveloped, and the country could not afford to lose the services of its largest bank. More research is needed to analyze the political forces that pressured the bank to retain its private banking functions – and the financial gains derived from these activities.